From mboxrd@z Thu Jan  1 00:00:00 1970
From: "Carlos R. Mafra" <crmafra@gmail.com>
Subject: Re: /proc/net/dev regression
Date: Sun, 11 Jan 2015 13:40:12 +0000
Message-ID: <20150111134012.GA2079@linux-g29b.site>
References: <20150110232518.GA3212@linux-g29b.site>
 <20150111002706.GC22149@ZenIV.linux.org.uk>
 <20150111010036.GD22149@ZenIV.linux.org.uk>
 <20150111013335.GA5753@linux-g29b.site>
 <20150111013913.GE22149@ZenIV.linux.org.uk>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: LKML <linux-kernel@vger.kernel.org>,
	Hauke Mehrtens <hauke@hauke-m.de>,
	"John W. Linville" <linville@tuxdriver.com>, netdev@vger.kernel.org
To: Al Viro <viro@ZenIV.linux.org.uk>
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <20150111013913.GE22149@ZenIV.linux.org.uk>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: netdev.vger.kernel.org

On Sun, 11 Jan 2015 at  1:39:13 +0000, Al Viro wrote:
> On Sun, Jan 11, 2015 at 01:33:35AM +0000, Carlos R. Mafra wrote:
> 
> > I think the problem with wmnet is not that it was expecting the fields
> > to be aligned because it never had problems before (when definitely more
> > than 10 megabytes were received, wmnet is crappy but not _that_ crappy).
> > 
> > I think the problem really was here,
> > 
> > 	totalbytes_in = strtoul(&buffer[7], NULL, 10);
> > 
> > After the patch the device name is 8 characters long and &buffer[7]
> > overlaps with the name instead of reading the bytes. Before the
> > patch is was fine because the call to strtoul() seems correct in the
> > sense that it would read everything until the NULL. So more than 10
> > megabytes was still ok.
> > 
> > So I guess I was wrong when suggesting that the problem was the
> > alignment.
> 
> Several lines below there's this:
>                         totalpackets_out = strtoul(&buffer[74], NULL, 10);
>                         if (totalpackets_out != lastpackets_out) {
>                                 totalbytes_out = strtoul(&buffer[66], NULL, 10);
>                                 diffpackets_out += totalpackets_out - lastpackets_out;
>                                 diffbytes_out += totalbytes_out - lastbytes_out;
>                                 lastpackets_out = totalpackets_out;
>                                 lastbytes_out = totalbytes_out;
>                                 tx = True;
>                         }
> 
> So I'm afraid it *is* that crappy.  This function really should use scanf();
> note that updateStats_ipchains() in the same file does just that (well,
> fgets()+sscanf() for fsck knows what reason).  And I'd be careful with all
> those %d, actually - it's not _that_ hard to get more than 4Gb sent.
> scanf formats really ought to match the kernel-side (seq_)printf one...

Ok, I fixed wmnet using Al's suggestion.

As far as I'm concerned, my regression complaint can be dismissed. It's
all working fine again.

Thanks!